198 BOTANY. Collomia lineaeis, Nutt (Watson, I. c. p. 261).—Utah, Watson's Rep.; South Park, Colorado, July, 1873, Wolf (686, 687). Collomu. geacilis, Dougl (Watson, I. c. p. 262).—Denver, Colo., June, 1873, Wolf (354). Collomia gilioides, Benth. (Gilia divaricata, Nutt Pl Gamb. p. 4; Watson, I. c. p. 464).—Willow Spring, Arizona, July, 1874, Rothrock (233). Collomia Cavanillesiana, Don. (Gray, I. c. p. 260; Watson, I. c. p. 465).—Big Dry Fork, Arizona, 1873, Loew (678); Camp Grant, at 5,250 feet elevation, Rothrock (439); Chiricahua Mountains, August 1, 1874, Rothrock (536); Black River, at 4,500 feet elevation, September, 1874, Rothrock (788). Collomia longifloba, Gray (Proc. Am. Acad. 8, p. 261; Watson, I. c. p. 465).—San Francisco Mountains, Arizona, 1871, 1872, Watson's Rep.; Denver, Colo., June, 1873, Wolf (675, 677) ; deserts of New Mexico, 1873, Loew; Camp Bowie, Aria., August, 1874, Rothiock (412); Deer Spring (180). Collomia aggeegata (Gilia aggregata, Spieng.; Watson, I. c. p. 269).— Arizona, Nevada, and Utah, 1871, 1872, Watson's Rep.; Arizona, 1873, Loew (305, 744); Blue River, Colorado, June, 1873, Wolf (745); Rocky Canon, Arizona, July, 1874, Rothrock (275) ; White Mountains, Arizona, at 9,000 feet elevation, September, 1874, Rothrock (810).—An examination of a goodly number of specimens of Gilia aggregata, Spreng., from different sta¬ tions in the Rocky Mountains and California, brings to light a character in the species which has been overlooked,—the unequal insertion of the stamens. The degrees of variation in this respect are remarkable and so gradual as to rule out the supposition of dimorphism. In the extreme cases, they are included, on the one side, in the tube of the corolla and very unequally inserted at points wide apart, from the base upward, whilst, on the other, they are crowded toward the throat and more or less exserted. Then, the inser¬ tion sometimes is, or rather, perhaps, appears to be, equal. In a specimen of this kind from California, it would be hard to declare that the filaments in some of the flowers do not start out from the same horizontal line, whilst in other floivers on the same stalk their insertion is decidedly unequal. Now, as the main character which separates Collomia from Gilia is the unequal inser-